Metal tolerance is often a result of metal storage or distribution. Thus, with the goal of advancing the molecular understanding of such metal homeostatic mechanisms, natural variation of metal tolerance in Arabidopsis thaliana was investigated. Substantial variation exists in tolerance of excess copper (Cu), zinc (Zn) and cadmium (Cd). Two accessions, Col-0 and Bur-0, and a recombinant inbred line (RIL) population derived from these parents were chosen for further analysis of Cd and Zn tolerance variation, which is evident at different plant ages in various experimental systems and appears to be genetically linked. Three QTLs, explaining in total nearly 50% of the variation in Cd tolerance, were mapped. The one obvious candidate gene in the mapped intervals, HMA3, is unlikely to contribute to the variation. In order to identify additional candidate genes the Cd responses of Col-0 and Bur-0 were compared at the transcriptome level. The sustained common Cd response of the two accessions was dominated by processes implicated in plant pathogen defense. Accession-specific differences suggested a more efficient activation of acclimative responses as underlying the higher Cd tolerance of Bur-0. The second hypothesis derived from the physiological characterization of the accessions is a reduced Cd accumulation in Bur-0.
Plant traits show enormous intraspecific natural variation. Molecular dissection of this variation can lead to new mechanistic insights into physiological or developmental processes and ideally yield an understanding of local adaptation to specific habitats. Most traits vary quantitatively and are controlled by multiple genes1. Arabidopsis thaliana represents an ideal system for molecular studies of natural variation because (i) it is distributed across a wide range of environments, and (ii) tremendous molecular knowledge and a vast array of genetic tools are available2. Accordingly, the natural variation in many different traits has already been studied in A. thaliana, enabling major progress in the understanding of diverse biological phenomena such as control of flowering time or herbivore defense3.
An important aspect of a plant’s environment is the composition of the soil. Plant roots are exposed to a huge range of as well as rapid fluctuations in the concentrations of bioavailable mineral ions. Relevant as environmental factors are not only macro- and microelements such as phosphorus or iron (Fe), but also potentially highly toxic elements without biological function, for example cadmium (Cd). Nonessential toxic elements are present in the environment either because of natural causes or because of man-made pollution. Cd has been released into the environment by industrial activities (e.g. metal smelting, battery manufacturing) and agricultural practices (e.g. use of phosphate fertilizers and sewage sludge)4, 5.
Plants have to acquire essential microelements (Fe, Zn, Cu, Mn, Co, Ni, Mo, B) from soil solutions that can vary in concentrations of the respective ions by orders of magnitude. Not only deficiency is a threat but also toxicity. An excess of, for instance, Cu or Zn ions can inhibit growth due to the tendency of these metal ions to strongly interact with various cellular components6. A homeostatic system comprising metal transporters, metal ligands and regulatory proteins maintains the concentrations of essential elements within rather narrow physiological ranges inside plant tissues and limits the accumulation of non-physiological elements7, 8. Variation in the so-called ionome, the mineral element composition of a tissue, an organ or an organism9, is the result of complex physiological processes and interactions with the environment10. Consequently, molecular dissection of ionome variation has led to a better functional understanding of metal transporters11, for example the Mo transporter MOT112 or the metal efflux protein FPN2/IREG213.
Natural variation in metal tolerance, i.e. the ability to withstand an excess of essential or nonessential metal ions, has received less attention, probably because metal excess is comparatively rare under natural conditions. However, many genes contributing to metal tolerance are part of the homeostatic network controlling metal accumulation and distribution. The Zn tolerance gene ZIF1 encodes a nicotianamine transporter and affects translocation of Zn and Fe14. The major plant metal tolerance mechanism, metal-activated synthesis of phytochelatins, has an influence on Zn mobility within plants15. Thus, analysis of metal tolerance variation can potentially improve the understanding of metal homeostasis and this has already been demonstrated. One of the earliest studies on variation in A. thaliana metal tolerance revealed a nonfunctional allele of the P-type ATPase gene HMA5 as the underlying cause of Cu hypersensitivity in the accession Cvi-016. The exploration of variation in Zn tolerance17 improved the functional understanding of FRD3, a transporter mediating the loading of metal chelating citrate into the xylem18. Variation in the recently identified arsenate (AsV) reductase ATQ1/HAC1 controls both differences in arsenate tolerance19 and the shoot accumulation of As20.
In search of novel plant metal tolerance/metal homeostasis factors we therefore explored natural variation in Zn, Cu and Cd tolerance. Our data confirmed the existence of substantial differences between A. thaliana accessions in metal tolerance and motivated a genetic analysis of Cd tolerance variation. We identified QTLs contributing to variation in the growth response to Cd exposure in a Col-0 × Bur-0 recombinant inbred line population. In addition we compared the transcriptome changes upon Cd exposure in these two parental accessions in order to find candidate genes for the major QTL on chromosome 5 and to probe the extent of variation in Cd responses between A. thaliana accessions.
Detection of natural variation in A. thaliana metal tolerance
We tested a panel of A. thaliana accessions, assembled to maximize genetic diversity (23 out of the Core 2421), for their tolerance against an excess of the micronutrients Zn and Cu as well as growth-inhibiting concentrations of Cd. Relative root growth (RRG) of metal-treated plants was used as readout to assess tolerance. We detected variation in Cd tolerance ranging from 31 ± 5 to 60 ± 12% RRG in the presence of Cd. Zn tolerance varied to a similar degree with extremes of 46 ± 17 and 82 ± 18% relative growth (Fig. 1). Huge variation was detected in Cu tolerance because of the extreme Cu-sensitivity of Cvi-0 (7.5 ± 3.4%) (Supplementary Fig. S1) which is attributable to a loss of HMA5 function16.
With the aim of dissecting the genetic basis for the observed natural variation in A. thaliana metal tolerance, we focused on a comparison of the accessions Col-0 and Bur-0. This choice was governed by the difference in Zn and Cd tolerance with Bur-0 showing better relative growth in the presence of both metals and, more importantly, the availability of a recombinant inbred line (RIL) population at the time22. First, Zn and Cd tolerance of the two parental accessions were assessed more closely. Both showed similar root growth in control conditions and a dose-dependent growth reduction when treated with Cd or excess Zn (Fig. 2). At 1.5 µM CdCl2 the difference in the RRG of Col-0 and Bur-0 became highly significant and largest differences were observed upon addition of 2 µM CdCl2 to the medium. Bur-0 reached 74 ± 20% of the length under control conditions (80 ± 24.1 mm) while Col-0 reached only 41 ± 7% of 84 ± 7.6 mm. Under excess Zn conditions the accessions showed significant differences at concentrations ranging from 40 to 140 µM, with highly significant differences in RRG at 80 µM. Here Bur-0 displayed a growth of 53 ± 14%. Col-0 reached only 35 ± 9% RRG.
Variation in metal tolerance observed for seedlings grown in vitro does not necessarily apply also to more natural metal excess conditions in soil. Therefore, Col-0 and Bur-0 were exemplarily analyzed in control and in Zn excess soil. Extractable metal content of the soil was analyzed via ICP-OES after HCl extraction. With Fe contents of 168 and 233 µg/g dry soil, Mn contents of 142 and 155 µg/g dry soil as well as Al contents of 267 and 294 µg/g dry soil, similar levels of metal ions could be detected in control and Zn-spiked soil. Zn levels were 52 and 756 µg/g dry soil in control and Zn-spiked soil, respectively. Growth was monitored over the course of the experiment by measuring leaf area (Fig. 3). In control soil the two accessions did not differ significantly. In Zn excess soil, however, Bur-0 showed much better growth than Col-0. After 7 d the relative leaf area (RLA) (% of control) was still at 98 ± 16% for Bur-0 while for Col-0 it had already dropped to 64 ± 4%. At the end of the experiment after 25 d the RLA of Bur-0 was reduced to 50 ± 7% while Col-0 plants showed an RLA of only 19 ± 1%, significantly less than Bur-0. Zn accumulation in leaves of plants grown in Zn excess soil was slightly lower in Bur-0 while Fe accumulation was strongly reduced compared to Col-0. Mn accumulation did not differ between the two accessions (Fig. 3c).
Final step in the phenotypic analysis of the contrasting accessions was the analysis of metal accumulation in hydroponically grown plants, i.e. under conditions that provide easy access to root tissue. For plants grown in medium with varying Cd concentrations lower Cd accumulation was consistently observed for Bur-0 relative to Col-0 in both roots and shoots (Fig. 4a). Zn differences between the accessions were less pronounced with Bur-0 showing slightly lower Zn levels in roots and higher levels in shoots at the highest tested Zn exposure (Fig. 4b). The micronutrients Fe and Mn were also assessed (Supplementary Fig. S2). Fe contents of Bur-0 roots were lower under all tested conditions. No other consistent differences in metal content between the accessions were found.
Genetic analysis of the difference in Cd tolerance between Col-0 and Bur-0
In order to assess whether the differences in Cd and Zn tolerance of seedlings are possibly linked, we initially analyzed 21 randomly selected RILs of the Bur-0 × Col-0 RIL population. The mean RRG under Zn and Cd stress conditions was strongly correlated (r = 0.7, P < 0.001) (Fig. 5a) indicating that differences in Cd and Zn tolerance between Col-0 and Bur-0 might be attributable to variation in the same genes. For the mapping of QTLs underlying natural variation in metal tolerance we then focused on Cd tolerance because the respective phenotypic difference between Col-0 and Bur-0 was more pronounced and also more robust than the difference in Zn tolerance (see Fig. 2).
Growth of the RIL population was assessed at 2 µM CdCl2 stress since the growth difference between Bur-0 and Col-0 was most prominent at this concentration (Fig. 2b). In 14 separate assays conducted to test the Cd tolerance of all RILs, Bur-0 and Col-0 showed a mean RRG of 65 ± 10% and 43 ± 10%, respectively. Of the 164 RILs 131 could be included in the mapping analysis. The germination frequency of the other 33 lines was too low to ensure accurate determination of the Cd tolerance. The RRG of the 131 selected lines was normally distributed and varied from 13.5 to 117.5% (Fig. 5b).
The program MapQTL5® was used to correlate genotypic and phenotypic information of all RILs. QTLs with a logarithm of odds (LOD) score of greater than 3 (95% confidence interval as calculated by MapQTL5®) were detected on chromosome 4 between 5.6 and 6.9 Mb as well as 13.2 and 17.7 Mb, and on chromosome 5 between 2.9 and 5.3 Mb (Fig. 6a,b). These QTLs explained 16, 12 and 23%, respectively, of the difference in Cd tolerance between Col-0 and Bur-0. Using the relative seedling weight as tolerance readout, QTL regions were detected in the same positions, albeit with less statistical significance (Supplementary Fig. S3). The mapped QTL regions contain 355, 299 and 759 genes. The only gene located within the QTLs that was previously implicated in Cd tolerance is HMA3 (At4G30120). This heavy metal ATPase was shown to influence natural variation in Cd accumulation in A. thaliana shoots23, and HMA3 overexpression resulted in a slight increase in Cd tolerance24. Col-0 is known to carry a defective HMA3 allele25. We therefore transformed Col-0 with the HMA3 Bur-0 allele. However, no consistent increase in Cd tolerance was observed when independent transgenic lines were compared to Col-0. The minor increase in Cd tolerance in one transgenic line did not correlate with the expression level of HMA3 Bur-0 (Supplementary Fig. S4), indicating that a functional HMA3 does not alleviate Cd toxicity under our assay conditions. Instead, other genes must be responsible for the observed difference in Cd tolerance between Bur-0 and Col-0. Even the phytochelatin-deficient Cd hypersensitive cad1-3 mutant in Col-0 background showed only a minor, non-significant gain in Cd tolerance when expressing the HMA3 Bur-0 allele (Supplementary Fig. S4).
Further analysis was focused on the QTL on chromosome 5 since it had the largest effect on the phenotypic difference between Bur-0 and Col-0. RILs containing mainly Bur-0 alleles in this QTL region were selected. Genetic markers were generated for the interval between 2.9 and 5.3 Mb and then used to narrow the QTL interval to between 2.9 and 3.84 Mb (Fig. 6c). Within this region, between At5g09370 and At5G11900, 292 genes are located. A comparison with a list of 243 A. thaliana genes annotated as metal homeostasis-related26 showed that none of these is positioned in this interval.
Identification of putative candidate genes
Because of this lack of obvious candidate genes we employed two different strategies to find the genes most plausibly involved in Cd tolerance. First, we searched for genes in the mapped interval that contain polymorphisms between the two accessions27, 28. SNPs and indels in coding regions were highlighted and those resulting in an amino acid exchange were prioritized. In 155 genes a total of 395 such SNPs were detected. They were analyzed for their putative impact on protein functionality. A SIFT analysis29, 30 revealed that 51 genes carry nonsynonymous SNPs or indels predicted to be deleterious for protein function (Supplementary Table S1). The Bur-0 genome apparently carries polymorphisms possibly rendering 49 of these proteins nonfunctional. Two proteins were predicted to be functional in Bur-0 but not in Col-0, including the Mg transporter MGT7 (At5g09690). One gene appears to be nonfunctional in both accessions (At5g09730).
In our initial screen we observed that Cvi-0 and St-0 showed levels of Zn and Cd tolerance similar to Bur-0 (Fig. 1). While QTLs are of course population-dependent we nonetheless analyzed all genes of the QTL region for their functionality in Cvi-0 and St-0 and compared the results to the SIFT analysis for Bur-0. This comparison showed that of the two genes which are predicted to be non-functional in Col-0 but functional in Bur-0, only MGT7 is also predicted to be functional in Cvi-0 and St-0 highlighting it as a candidate for the observed QTL. Furthermore 11 of the 49 genes, predicted to be non-functional in Bur-0, are also predicted as non-functional in Cvi-0 and St-0 (Supplementary Table S2).
The second strategy searched for expression differences between Col-0 and Bur-0 in genes present in the mapped interval. A microarray experiment was conducted to find Cd responsive genes and to search for differential expression in Bur-0 and Col-0. Because this analysis is genome-wide it additionally provided information on natural variation of transcriptional responses to Cd exposure in A. thaliana. A comparatively mild Cd treatment was chosen in order to avoid confounding effects through differences in the severity of stress for the two accessions. Hydroponic growth in medium with 1 µM CdCl2 resulted in comparable growth reduction as determined by RLA: 57 ± 11% for Bur-0 and 58 ± 5% for Col-0 (Supplementary Fig. S5). As a sensitive biochemical marker for Cd exposure we analyzed the accumulation of PCs. PC synthesis is activated by cytosolic Cd and therefore can be used as readout for biologically active symplastic Cd levels. Comparable amounts of PC2 and PC3 accumulated in roots and shoots of Col-0 and Bur-0 during the Cd exposure indicating equal activation of AtPCS1 (Supplementary Fig. S5). Transcriptome changes in roots as the tissue of immediate Cd impact were analyzed using the ATH1 chip. For 183 genes in the QTL interval expression was detected. Out of these 35 genes showed differences between Col-0 and Bur-0. They either responded only in Bur-0 or only in Col-0, or were responsive in both accessions but showed at least a twofold difference in transcript level between the accessions under control or treatment conditions (Supplementary Fig. S6). For example, At5g09570 encoding a Cox19-like protein of unknown function showed higher expression level under control conditions and was more Cd responsive in Bur-0. Transcript abundance of the putative oligopeptide transporter gene At5g11570 was Cd-repressed only in Col-0.
Natural variation in A. thaliana Cd responses
The transcriptome data enabled global analysis of variation in Cd response. In total the expression of 16,022 genes was analyzed in root tissue. Fold changes were determined for each gene after Cd treatment and expression was compared between Bur-0 and Col-0. A differential expression was defined as a ≥2fold change of expression and a P adj of <0.05. In Col-0 698 genes were up-regulated upon Cd treatment and 736 were down-regulated. In Bur-0 more genes were Cd responsive, namely 1090 were up- and 722 were down-regulated. 522 and 388 genes, respectively, responded qualitatively in the same way in both accessions (Supplementary dataset S1). An enrichment analysis of the functional classes of those 2336 genes was performed and significant over- or underrepresentation determined (Fig. 7 and Supplementary dataset S2). In this way an A. thaliana Cd core response can be defined. Several biological processes activated in both accessions (left column in Fig. 7) are related to immune responses, e.g. “systemic acquired resistance”, “induced systemic resistance”, “response to jasmonic acid”, “response to salicylic acid”, “defense response to bacterium, incompatible interaction”. Part of the Cd core response was also the up-regulation of abiotic stress-associated processes such as “cellular response to hypoxia”, “response to oxidative stress” and “response to toxic substance”. Furthermore, several genes belonging to the GO category “sulfur compound metabolic process” were more active in both Col-0 and Bur-0 upon Cd exposure. A common theme of GO terms enriched among down-regulated genes was the cell wall: “cell wall biogenesis”, plant-type cell wall loosening”, and “xyloglucan metabolic process”. Also down-regulated was “glucosinolate biosynthetic process”. Several GO terms were specifically enriched among the genes responsive only in Bur-0. Most of them represent additional biological processes involved in pathogen defense, e.g. “innate immune response”, “response to chitin”, “response to fungus”. In contrast, no GO terms were enriched in the group of genes specifically up-regulated in Col-0.
Cd as an environmental pollutant is almost ubiquitously present at the global scale due to diffuse pollution over centuries. Important sources include metal smelting and the application of phosphate fertilizers with Cd impurities4. Humans are chronically exposed to low levels of Cd mostly through the consumption of plant-derived food because plants tend to take up and accumulate Cd5. Recent risk assessments, for instance by the European Food Safety Authority (EFSA), concluded that (i) the provisional tolerable weekly intake of Cd should be lowered to reduce health risks such as renal damage, osteoporosis, cardiovascular disease, and that (ii) current levels of human exposure to Cd are on average around this suggested lowered threshold of 2.5 µg/kg body weight (b.w.). Therefore, environmental exposure to Cd needs to be reduced (EFSA 2009, 2012). An efficient way of achieving this would be the development of crops with strongly reduced Cd accumulation rates, which in turn would benefit from a mechanistic understanding of genes and proteins mediating the movement of Cd within a plant. Because accumulation rates of metals and the ability to cope with an excess of the same metals are often tightly linked, e.g. in the case of the arsenate reductase ATQ1/HAC119, 20, the elucidation of tolerance mechanisms represents one way of identifying genes involved in the handling of Cd by plants.
Forward genetic screens were instrumental in elucidating the molecular basis of plant Cd tolerance mechanisms. The analysis of A. thaliana cad mutants demonstrated the key role of the phytochelatin pathway for Cd detoxification. Mutants with defects in glutathione or phytochelatin biosynthesis are strongly Cd hypersensitive31. We aimed at finding new factors contributing to Cd tolerance by studying natural variation within A. thaliana.
While some early reports on metal tolerance variation could not be confirmed32 we found within a core set of genetically diverse A. thaliana accessions21 about twofold variation for Zn and Cd tolerance at the seedling stage. In the case of Zn this value is between the variation factors of 1.6 and 4.4 reported for the highest and the lowest growth inhibitory Zn concentrations, respectively, applied to a very similar set of A. thaliana accessions17. More in-depth analyses of two accessions, Col-0 and Bur-0, indicated that differences in Zn tolerance are far stronger when plants are cultivated in Zn excess soil (Fig. 3). Bur-0 showed much better growth than Col-0. Corresponding experiments with Cd were not performed for safety reasons as the addition of growth-inhibitory Cd concentrations to soil could result in the exposure of experimenters to Cd-containing dust. Because Cd and Zn tolerance variation appear to be linked (Fig. 5a), the analysis of Cd tolerance variation may well reveal hidden factors of Zn homeostasis as was discussed for HMA3. The existence of functional and nonfunctional HMA3 alleles explains variation in leaf Cd accumulation. There is also a contribution to Zn accumulation differences but this is much weaker, most likely because Zn accumulation is influenced by more genes than Cd accumulation23.
A possible explanation of higher Bur-0 Cd tolerance is reduced Cd accumulation relative to Col-0. In hydroponically grown plants root and shoot Cd contents were consistently lower in Bur-0 (Fig. 4), again illustrating the connection between tolerance and accumulation. For Zn the situation is less clear. In Zn excess soil Bur-0 leaves contained slightly less Zn than Col-0. On the other hand, the situation in hydroponic culture is less clear with trends depending on tissue and metal concentration. In leaves Bur-0 accumulated more Zn while in roots a lower Zn accumulation was visible at the highest concentration the plants were exposed to. Data from the IonomicsHub (Baxter 2007) suggest that leaves of soil-grown Bur-0 plants accumulate less Cd than Col-0, which would be consistent with findings reported here. Publicly available data for two trays (868, 1060) with Col-0 and Bur-0 showed 30% and 9% less Cd for Bur-0 (ionomicshub. org). Zn content varied much less between the accessions (0 to 12% in trays 868, 1060, 1301, and 1482), again consistent with our observations.
The most extreme phenotypic difference found between Col-0 and Bur-0 in our study, namely the dramatically better leaf growth of Bur-0 on Zn excess soil, may at least in part be due to the higher Fe accumulation of Col-0 under these conditions and thus Fe toxicity (Fig. 3c). However, the mechanistic basis for this difference in Fe accumulation is unclear.
Analysis of Cd tolerance across the Bur-0 × Col-0 RIL population revealed strong variation for this trait (Fig. 5b). Normal distribution indicated the contribution of several genetic factors to the variation as was observed for two other RIL populations, Ler-0 × Col-4 and Col-gl1 × Kas-1, before33. Three QTLs could be mapped, the major one on chromosome 5 explaining 23% of the variation. Not surprisingly, the mapped intervals do not overlap with the previously mapped QTLs regions contributing to variation in the above-mentioned RIL populations33.
A large number of QTLs involved in abiotic stress tolerance have been detected in A. thaliana to date. However, rather few have been molecularly elucidated34. As was the case here, the mapped intervals are usually several Mb wide. Molecular elucidation of QTLs is therefore largely dependent on whether or not genes that have previously been implicated in the trait in question are located in the mapped interval. Examples of successful identification of the causal polymorphisms include APR2 for sulfate content35, MOT1 for Mo accumulation12, and FRD3 for Zn tolerance18. Within the three mapped Cd tolerance QTLs only HMA3 had been implicated in Cd accumulation23 and tolerance24 before. The latter was not confirmed by the data presented here. Transformation of Col-0 or even the Cd hypersensitive mutant cad1-3, which both lack a functional HMA3 allele due to a deletion of one nucleotide in exon 8, with the Bur-0 allele of HMA3 under control of its own promotor did not enhance Cd tolerance. This discrepancy is possibly attributable to expression differences as the previously reported effect of HMA3 on Cd tolerance was conferred by constitutive overexpression. Overall, the proposed contribution of HMA3 polymorphisms to variation in A. thaliana Cd tolerance24 appears to be absent across the Bur-0 × Col-0 RIL population.
Further fine-mapping of the major QTL on chromosome 5 using an F2 population derived from a cross of RIL 243 with Col-0 failed because we did not find a sufficient number of recombination events in the Bur-0 region spanning the QTL. Thus, in the absence of other obvious candidate genes we attempted to find potential candidate genes (i) by evaluating polymorphisms in the mapped region between At5g09350 and At5g11950, and (ii) by comparing the transcriptional responses of Col-0 and Bur-0 to Cd exposure. The former strategy was based on the analysis of nonsynonymous SNPs and indels. It suggested the Mg transporter genes AT5G09690 (MGT7/MRS2-7) and AT5G09720 for further study given the known effect of Mg status on Cd tolerance36. The status of MGT7 as a candidate gene is further supported by the SIFT analyses for the accessions Cvi-0 and St-0 (Supplementary Table S2) which showed Bur-0-like growth responses to Cd and Zn toxicity.
The analysis of global transcriptome changes in the roots of Col-0 and Bur-0 was carried out with older, hydroponically grown plants. This ensured access to sufficient root material and allowed comparative assessment of the degree of Cd stress on the two accessions. Accumulation of PCs is a suitable indicator for intracellular, biologically active Cd because PC synthesis, which is mostly attributable to AtPCS1, is strongly activated by the presence of Cd. Levels of the two major phytochelatins, PC2 and PC3, were similar for roots and shoots of Col-0 and Bur-0 (Supplementary Fig. S5). This demonstrated comparable stress levels. Also, in order to avoid variation caused by slight variation in kinetics of the Cd responses we analyzed a comparatively late time point, i.e. after one week of exposure. Within the QTL region on chromosome 5 several genes were found to be responsive to Cd. These included the sulfate transporter SULTR2;1 gene (At5g10180) and At5g10380 (encoding a putative E3 ligase) which were previously found to be robustly Cd-responsive in roots at earlier time points (2, 6, and 30 h)37. However, based on current knowledge, no obvious new candidate genes emerged from the transcript analysis of genes in the chromosome 5 interval.
A genome-wide comparison of the Col-0 and Bur-0 root Cd responses can help define transcriptome changes conserved in A. thaliana. The focus of our analysis was on sustained acclimative changes. This is complementary to available data for much earlier time points between 2 h and 30 h (e.g. refs 37,38,39). Nonetheless substantial overlap between the core response shown by Col-0 and Bur-0 and more rapid transcriptome changes is apparent. Among the 540 up-regulated and 400 down-regulated genes constituting the Cd core response reported here, 36 and 39%, respectively, showed corresponding changes in roots of 4 week old hydroponically grown A. thaliana Col-0 plants exposed to Cd2+ for 30 h37. Similarly, 38% of the genes found to be up-regulated in roots of 10 d old A. thaliana Col-4 seedlings exposed for 24 to Cd2+ 40 are on the list of common Cd-induced genes in Col-0 and Bur-0. Thus, a considerable fraction of the acclimative adjustments in Cd exposed roots appears to be long-lasting. The most conspicuous pattern apparent from the GO term enrichment analysis is the dominance of gene groups associated with plant defense. Several of the overrepresented GO terms among core up-regulated genes belong into the immunity context. The same is true when the analysis is restricted to the genes up-regulated both after 30 h37 and after 7 d (data reported here). Furthermore, this observation is in accordance with the finding that when different toxic ions (Al3+, Cu2+, Cd2+, Na+) were analyzed for their effect on the A. thaliana root transcriptome, defense genes were enriched in the group of genes specifically responsive to Cd stress40. The functional significance of this congruence in transcriptome changes remains unclear. Possibly the type of oxidative stress caused by Cd exposure41 bears similarities with the elevated ROS production typical for immune responses. Alternatively, it might be more than an oddity that there are several proteins implicated in both metal tolerance and plant nonhost resistance, e.g. ABCG36/PDR8/PEN342, 43 and AtPCS144.
Apparently not sustained over longer exposure times is one of the major acclimations of plants to Cd stress, namely the up-regulation of sulfur assimilation. There is ample evidence from transcriptome, proteome and metabolome studies that in order to replenish glutathione consumed for phytochelatin synthesis45, sulfate uptake and activation, the reduction of adenosine phosphosulfate as well as the synthesis of cysteine are strongly and rapidly activated39, 40, 46, 47. However, similar to the findings of van de Mortel et al.48 for A. thaliana roots exposed to Cd for one week as well, sulfur assimilation genes are not enriched among the responsive genes. While single members of the SULTR sulfate transporter family (e.g. SULTR2;1) are still up-regulated after 7 d, the enrichment of the GO term “sulfur compound metabolic process” (Fig. 7) is attributable rather to genes encoding glutathione S-transferases. Another sulfur metabolism-related long-term change is the down-regulation of glucosinolate biosynthesis. This has been hypothesized as a mechanism to make more sulfur available for PC synthesis37 and according to our data indeed represents a sustained core response of Cd-exposed A. thaliana roots.
The analysis of accession-specific responses suggested a second possible explanation for the higher Cd tolerance of Bur-0 besides the reduced Cd accumulation (Fig. 4a). Bur-0 roots responded more strongly to Cd than Col-0 even though according to the PC accumulation data there was no difference in the degree of effective Cd exposure, i.e. available Cd in the cytosol. The stronger response is not only apparent in the larger number of up- and down-regulated genes. Many of the core response GO terms as well as several other related GO terms, mainly additional defense-associated biological processes such as “response to chitin”, “innate immune response” and “defense response to bacterium”, were enriched among the genes responsive to Cd only in Bur-0. Thus, a possible explanation for the higher Cd tolerance of Bur-0 could be a more efficient activation of acclimative transcriptome changes. The understanding of signal transduction events mediating Cd responses is still fragmentary. Recently, the transcription factor ZAT6 was assigned a role in the activation of genes involved in the phytochelatin pathway49. Overexpression of ZAT6 conferred higher Cd tolerance. Interestingly, ZAT6 and one of its target genes (ABCG36/PDR8/PEN3) as well as many other genes associated with Cd tolerance, e.g. SULTR3;4, SULTR4;1, SBP150, ABCC251, and ABCC352 were up-regulated specifically in Bur-0. Thus, there could be a difference between Bur-0 and Col-0 in the events upstream of ZAT6 activation. Responsible genes in the mapped Cd tolerance QTLs will hopefully be identified when additional information becomes available on gene functions and on chromosomal regions influencing variation in Cd tolerance across other A. thaliana populations.
Arabidopsis thaliana lines
A. thaliana accessions were obtained from the Nottingham Arabidopsis Stock Center (NASC). The Col-0 × Bur-0 RIL population was obtained from the Versailles Arabidopsis Stock Center and was developed by Simon et al.22. The core population of 164 RILs was used for the mapping-related phenotyping.
Plants were cultivated in 1/10 Hoagland medium under long day conditions of 16 h light (50 µE) /8 h dark in a cultivation room. For the assessment of metal tolerance phenotypes a medium solidified with 1% (w/v) agarose (Type A, Sigma Aldrich) was used. The nutrient solution contained 100 µM (NH4)2HPO4, 200 µM MgSO4, 280 µM Ca(NO3)2, 600 µM KNO3, 5 µM Fe-HBED and 1% (w/v) sucrose and was buffered with 5 mM MES at pH 5.7. Metal salts were added as indicated.
For hydroponic culture plants were grown in a modified, liquid 1/10 Hoagland medium (87.1 µM (NH4)2HPO4, 200 µM MgSO4, 400 µM Ca(NO3)2, 600 µM KNO3, 5 µM Fe-HBED, 4.63 µM H3BO3, 0.032 µM CuSO4, 0.915 µM MnCl2, 0.077 µM ZnSO4, 0.011 µM MoO3; pH 5.7). Starting at week 2 of the plant cultivation the Bur-0 and Col-0 individuals were kept in 50 ml Falcon tubes. During the 3rd and 4th week the medium was exchanged once per week. From week 5 on the medium was exchanged twice per week. The plants were cultivated in climate chambers at short day conditions of 8 h light (200 µE)/16 h dark with 22 °C during the light period and 20 °C during the dark period.
Experiments were performed as described previously53. Control soil was spiked with 400 mg ZnSO4 per kg soil (dry weight) and Zn contaminated soil was prepared by adding 6400 mg ZnSO4 per kg soil.
In order to determine the Cd tolerance of hydroponically grown plants, the leaf growth was monitored by determining leaf area over the course of the cultivation. For the quantification of the leaf area, pictures were taken (same distance from the objects and a marker of known size) and analyzed (Adobe Photoshop CS2, version 9.0). Pixels of similar color were selected by use of the “Magic Wand Tool” at a tolerance of 32 using the neighboring and smoothing tool. Pixels of the leaf area were normalized to the pixels covering a standard of known size.
164 RILs were phenotyped under control conditions and in the presence of 2 µM CdCl2. The relative root growth (RRG) was calculated for each RIL as the mean root length on Cd plates normalized to the mean root length on control plates. RRG and genotypic information22 were loaded into the program MapQTL5®54. Standard methods for interval mapping55 were applied and the LOD score for each position on the genome was determined as an output for the likelihood of the association of a particular genomic region with the Cd tolerance phenotype.
PCR based marker analysis
For further genotyping of interesting RILs, gDNA was extracted from young leaves or seedlings. To 100 mg pulverized, frozen plant material 500 µl extraction buffer (200 mM Tris/HCl pH 7.5, 250 mM NaCl, 25 mM EDTA, 0.5% SDS) were added. The DNA contained in the aqueous phase was then precipitated using isopropanol (1:1 ratio of DNA fraction to ethanol) and washed with ethanol. After resuspension in 50 µl deionized water, gDNA was used for PCR amplification. PCR was performed using the following protocol: 94 °C 1 min initial denaturation followed by 35 cycles of 94 °C 30 sec; X °C 30 sec; 72 °C 1 min/1 kb followed by a final elongation step at 72 °C for 3 min. The annealing temperature X was chosen for each primer pair 2 °C below the lower Tm of the two primers. Four CAPS markers were generated between the markers *2900 and *4011 (*3109 fw: TTTCCTTACAAGCCTGAAATATCC, rev: AAACACATCGCACATTTTGAAC, flanking a BseLI restriction site specific for Col-0; *3343 fw: GATGCCATAGAGATGAATGCAG, rev: ATGGGCCTATGTGTAATCAACC, flanking a TaqI restriction site specific for Bur-0; *3558 fw: TTCCTCAGATCATCCATGTGTC, rev: TGCAAATACTAATGCTCCAACG, flanking a KpnI restriction site specific for Col-0; *3782 fw: TGTTGTGCCTTGATTTTCAGTC, rev: TGGGTGCAAATTTCTCCTCTAC, flanking a TasI restriction site specific for Col-0) and 17 RILs were analyzed for their genotype in this region.
Plants were grown hydroponically. After one week of Cd exposure, roots and shoots of 3-4 individuals were pooled and two pools for each treatment and accession were generated. Harvested plant material was immediately frozen in liquid nitrogen. Plants were harvested early in the morning at the beginning of the light period. RNA from root material was extracted by adding 1 ml of TRIzol® (Invitrogen) to 100 mg of frozen plant powder and, after mixing, adding 200 µl chloroform. From the resulting aqueous phase the RNA was precipitated using isopropanol in a 1:1 ratio. The washed RNA was resuspended in 20 µl RNase-free H2O. Further purification was achieved by using the RNaeasy Kit from Quiagen, following the instruction manual of the manufacturer.
For the microarray analysis the extracted RNA was translated into aRNA with the GeneChip® 3′ IVT Express Kit from Affymetrix. Fragmented aRNA was hybridized to an Arabidopsis GeneChip microarray56 at the „Zentum für medizinische Grundlagenforschung der Martin-Luther-Universität“ in Halle/Saale, Germany. An Affymetrix GeneChip Scanner 3000 was used to analyze the hybridized microarray, applying an Affymetrix GeneChip Operating Software (GCOS). Program R (version 3.0.0)57 and the Bioconductor packages58 were used for data normalization, pre-processing and quality control. Signals were normalized using the Robust Multi-chip Average (RMA) method at standard settings. A linear model was used to calculate fold changes and P adj. values for each gene. The lmFit function was used with standard settings. The affy package was employed for the analysis of the data, quality control and annotation. Differentially expressed genes were selected based on statistically significant (P adj ≤ 0.05) expression differences of at least 2fold between the genotypes under equal conditions or between treatments within each genotype. GO enrichment was performed with the PANTHER classification system (database release 27.10.2016) developed by the Gene Ontology Consortium (Mi et al., 2013). Cluster analysis was performed using the software R (version 3.3.1). Raw data of the microarray experiment reported here can be accessed via the GEO (Gene Expression Omnibus) data base under the accession number GSE94314.
For quantitative real-time (qRT) analysis of HMA3 transcript abundance cDNA was synthesized from DNaseI treated RNA extracted from seedlings, using the RevertAid First Strand cDNA Synthesis Kit (Thermo-Fisher). The qRT PCR was performed using SYBR Green (iQ SYBR Green supermix, Bio-Rad) in a 96-well iCycler with a MyiQ real-time RCR detection system. The master mix contained 5 pmol of forward and reverse primer (HMA3_Bur_fw: CCAAACTGCAATGCTCACAG; HMA3_Bur_rev: TGGCCCTTGGATCTTGAAATC) and the PCR product was amplified from 5 µl of a 1:50 dilution of cDNA in a total volume of 20 µl. The PCR program was 95 °C for 10 min, 40 cycles of 95 °C for 10 sec and 60 °C for 1 min. The relative transcript level was determined as 1000 × 2−ΔCT with the cycle threshold difference ΔCT being CTTarget gene − CTEF1α.
Insertions and deletions in the Bur-0 genome within the QTL interval on chromosome 5 were downloaded from the 1001 Genomes Data Center: (http://1001genomes.org/data/MPI/MPIOssowski2008/releases/current/strains/Bur-0/). Sequence differences between Col-0 and Bur-0 were validated using the 1001 genomes browser: http://signal.salk.edu/atg1001/3.0/gebrowser.php. The most recent information about single nucleotide polymorphisms (SNPs) between Col-0 and Bur-0, stemming from the resequencing efforts of the 1001 genomes project, were downloaded from the GWA-Portal: https://gwas.gmi.oeaw.ac.at/#/genotype/snpviewer. Line 7058 was chosen to extrapolate all SNP positions leading to amino acid differences in the protein sequences between Col-0 and Bur-0 within the QTL region on chromosome 5. Line 7058 was chosen since it was the line used in the resequencing efforts and has the most SNPs annotated. SNPs were used to predict potentially non-functional AA exchanges in Bur-0 using the SIFT4G application (Vaser et al., 2016).
Phytochelatins (PCs) were determined by UPLC-ESI-QTOF-MS as described earlier53. Briefly, frozen homogenized plant material was extracted with 0.1% (v/v) trifluoroacetic acid containing 6.3 mM diethylene triamine pentaacetic acid and 40.04 µM N-acetylcysteine as internal standard. Prior to derivatization with monobromobimane at 45 °C for 30 min, thiol groups were reduced with Tris-(2-carboxyethyl)-phosphine. Labelled thiols were separated on a HSS T3 column (1.8 μm, 2.1 × 100 mm; Waters Corporation, Milford, MA, USA) by a Waters Aquity UPLC system applying a linear binary gradient of water (A) and acetonitrile (B), both acidified with 0.1% (v/v) formic acid, at a flow of 0.5 mL min–1: 99.5% A, 0.5% B for 1 min, a linear gradient to 60.5% B at 10 min, gradient to 99.5% B at 12 min, flushing with 99.5% B for 1 min, a gradient back to initial conditions in 1 min and an additional re-equilibration for 1 min. The column temperature was set to 40 °C. Thiols were detected with a Q-TOF Premier mass spectrometer equipped with an ESI-source (Waters Corporation) operated in the V + mode. For quantification the QuanLynx module of the MarkerLynx software was used.
Inductively Coupled Plasma-Optical Emission Spectroscopy
Dry plant material was, after washing at 4 °C with H2O (twice, 10 min), 100 mM CaCl2 (once, 10 min), 50 mM EDTA (once, 5 min), digested in 4 ml HNO3 and 2 ml H2O2 for 12 min at 180 °C using a microwave (Start 1500, MLS GmbH). Metal concentrations were measured via ICP-OES (ICAP 6500, Thermo): Zn at 213.8 nm, Fe at 238.2 nm, Cd at 226.5 nm and Mn at 257.6 nm.
Alonso-Blanco, C. et al. What has natural variation taught us about plant development, physiology, and adaptation? Plant Cell 21, 1877–1896, doi:10.1105/tpc.109.068114 (2009).
Mitchell-Olds, T. & Schmitt, J. Genetic mechanisms and evolutionary significance of natural variation in Arabidopsis. Nature 441, 947–952, doi:10.1038/nature04878 (2006).
Weigel, D. Natural variation in Arabidopsis: From molecular genetics to ecological genomics. Plant Physiol. 158, 2–22, doi:10.1104/pp.111.189845 (2012).
UNEP. Draft final review of scientific information on cadmium (2008).
Clemens, S., Aarts, M. G. M., Thomine, S. & Verbruggen, N. Plant science: the key to preventing slow cadmium poisoning. Trends Plant Sci. 18, 92–99, doi:10.1016/j.tplants.2012.08.003 (2013).
Lin, Y.-F. & Aarts, M. G. M. The molecular mechanism of zinc and cadmium stress response in plants. Cell. Mol. Life Sci. 69, 3187–3206, doi:10.1007/s00018-012-1089-z (2012).
Clemens, S. Molecular mechanisms of plant metal tolerance and homeostasis. Planta 212, 475–486, doi:10.1007/s004250000458 (2001).
Palmer, C. M. & Guerinot, M. L. Facing the challenges of Cu, Fe and Zn homeostasis in plants. Nat. Chem. Biol. 5, 333–340, doi:10.1038/nchembio.166 (2009).
Salt, D. E., Baxter, I. & Lahner, B. Ionomics and the study of the plant ionome. Annu. Rev. Plant Biol. 59, 709–733, doi:10.1146/annurev.arplant.59.032607.092942 (2008).
Baxter, I. & Dilkes, B. P. Elemental profiles reflect plant adaptations to the environment. Science 336, 1661–1663, doi:10.1126/science.1219992 (2012).
Huang, X.-Y. & Salt, D. E. Plant ionomics: From elemental profiling to environmental adaptation. Mol. Plant 9, 787–797, doi:10.1016/j.molp.2016.05.003 (2016).
Baxter, I. et al. Variation in molybdenum content across broadly distributed populations of Arabidopsis thaliana is controlled by a mitochondrial molybdenum transporter (MOT1). Plos Genet. 4 (2008).
Morrissey, J. et al. The ferroportin metal efflux proteins function in iron and cobalt homeostasis in Arabidopsis. Plant Cell 21, 3326–3338, doi:10.1105/tpc.109.069401 (2009).
Haydon, M. J. et al. Vacuolar nicotianamine has critical and distinct roles under iron deficiency and for zinc sequestration in Arabidopsis. Plant Cell 24, 724–737, doi:10.1105/tpc.111.095042 (2012).
Kühnlenz, T. et al. Phytochelatin synthesis promotes leaf Zn accumulation of Arabidopsis thaliana plants grown in soil with adequate Zn supply and is essential for survival on Zn-contaminated soil. Plant Cell Physiol. 57, 2342–2352, doi:10.1093/pcp/pcw148 (2016).
Kobayashi, Y. et al. Amino acid polymorphisms in strictly conserved domains of a P-type ATPase HMA5 are involved in the mechanism of copper tolerance variation in Arabidopsis. Plant Physiol. 148, 969–980, doi:10.1104/pp.108.119933 (2008).
Richard, O. et al. Diversity analysis of the response to Zn within the Arabidopsis thaliana species revealed a low contribution of Zn translocation to Zn tolerance and a new role for Zn in lateral root development. Plant Cell Environ. 34, 1065–1078, doi:10.1111/j.1365-3040.2011.02305.x (2011).
Pineau, C. et al. Natural variation at the FRD3 MATE transporter locus reveals cross-talk between Fe homeostasis and Zn tolerance in Arabidopsis thaliana. PLoS Genet. 8, e1003120, doi:10.1371/journal.pgen.1003120 (2012).
Sánchez-Bermejo, E. et al. Natural variation in arsenate tolerance identifies an arsenate reductase in Arabidopsis thaliana. Nat. Commun. 5, 4617, doi:10.1038/ncomms5617 (2014).
Chao, D.-Y. et al. Genome-wide association mapping identifies a new arsenate reductase enzyme critical for limiting arsenic accumulation in plants. PLoS Biol. 12, e1002009, doi:10.1371/journal.pbio.1002009 (2014).
McKhann, H. I. et al. Nested core collections maximizing genetic diversity in Arabidopsis thaliana. Plant J. 38, 193–202, doi:10.1111/tpj.2004.38.issue-1 (2004).
Simon, M. et al. Quantitative trait loci mapping in five new large recombinant inbred line populations of Arabidopsis thaliana genotyped with consensus single-nucleotide polymorphism markers. Genetics 178, 2253–2264, doi:10.1534/genetics.107.083899 (2008).
Chao, D.-Y. et al. Genome-wide association studies identify heavy metal ATPase3 as the primary determinant of natural variation in leaf cadmium in Arabidopsis thaliana. PLoS Genet. 8, e1002923, doi:10.1371/journal.pgen.1002923 (2012).
Morel, M. et al. AtHMA3, a P-1b-ATPase allowing Cd/Zn/Co/Pb vacuolar storage in Arabidopsis. Plant Physiol. 149, 894–904, doi:10.1104/pp.108.130294 (2009).
Hussain, D. et al. P-type ATPase heavy metal transporters with roles in essential zinc homeostasis in Arabidopsis. Plant Cell 16, 1327–39, doi:10.1105/tpc.020487 (2004).
Mandáková, T., Singh, V., Krämer, U. & Lysak, M. A. Genome structure of the heavy metal hyperaccumator Noccaea caerulescens and its stability on metalliferous and non-metalliferous soils. Plant Physiol, doi:10.1104/pp.15.00619 (2015).
Ossowski, S. et al. Sequencing of natural strains of Arabidopsis thaliana with short reads. Genome Res. 18, 2024–2033, doi:10.1101/gr.080200.108 (2008).
Seren, Ü. et al. GWAPP: A web application for genome-wide association mapping in Arabidopsis. Plant Cell 24, 4793–4805, doi:10.1105/tpc.112.108068 (2012).
Kumar, P., Henikoff, S. & Ng, P. C. Predicting the effects of coding non-synonymous variants on protein function using the SIFT algorithm. Nat. Protoc. 4, 1073–1081, doi:10.1038/nprot.2009.86 (2009).
Ng, P. C. & Henikoff, S. SIFT: predicting amino acid changes that affect protein function. Nucleic Acids Res. 31, 3812–3814, doi:10.1093/nar/gkg509 (2003).
Cobbett, C. & Goldsbrough, P. Phytochelatins and metallothioneins: Roles in heavy metal detoxification and homeostasis. Annu. Rev. Plant Physiol. Plant Mol. Biol. 53, 159–182, doi:10.1146/annurev.arplant.53.100301.135154 (2002).
Silva-Guzman, M., Addo-Quaye, C. & Dilkes, B. P. Re-evaluation of reportedly metal tolerant Arabidopsis thaliana accessions. PloS One 11, e0130679, doi:10.1371/journal.pone.0130679 (2016).
Tazib, T. et al. Quantitative trait loci controlling resistance to cadmium rhizotoxicity in two recombinant inbred populations of Arabidopsis thaliana are partially shared by those for hydrogen peroxide resistance. Physiol. Plant. 136, 395–406, doi:10.1111/ppl.2009.136.issue-4 (2009).
Assmann, S. M. N. Natural variation in abiotic stress and climate change responses in Arabidopsis: Implications for twenty-first-century agriculture. Int. J. Plant Sci. 174, 3–26, doi:10.1086/667798 (2013).
Loudet, O. et al. Natural variation for sulfate content in Arabidopsis thaliana is highly controlled by APR2. Nat. Genet. 39, 896–900, doi:10.1038/ng2050 (2007).
Hermans, C., Chen, J., Coppens, F., Inzé, D. & Verbruggen, N. Low magnesium status in plants enhances tolerance to cadmium exposure. New Phytol. 192, 428–436, doi:10.1111/nph.2011.192.issue-2 (2011).
Herbette, S. et al. Genome-wide transcriptome profiling of the early cadmium response of Arabidopsis roots and shoots. Biochimie 88, 1751–1765, doi:10.1016/j.biochi.2006.04.018 (2006).
Weber, M., Trampczynska, A. & Clemens, S. Comparative transcriptome analysis of toxic metal responses in Arabidopsis thaliana and the Cd2+-hypertolerant facultative metallophyte Arabidopsis halleri. Plant Cell Environ. 29, 950–963, doi:10.1111/pce.2006.29.issue-5 (2006).
Jobe, T. O. et al. Feedback inhibition by thiols outranks glutathione depletion: a luciferase-based screen reveals glutathione-deficient γ-ECS and glutathione synthetase mutants impaired in cadmium-induced sulfate assimilation. Plant J. 70, 783–795, doi:10.1111/j.1365-313X.2012.04924.x (2012).
Zhao, C.-R. et al. Comparative transcriptomic characterization of aluminum, sodium chloride, cadmium and copper rhizotoxicities in Arabidopsis thaliana. BMC Plant Biol. 9, 32, doi:10.1186/1471-2229-9-32 (2009).
Cuypers, A. et al. Cadmium stress: an oxidative challenge. Biometals 23, 927–940, doi:10.1007/s10534-010-9329-x (2010).
Stein, M. et al. Arabidopsis PEN3/PDR8, an ATP binding cassette transporter, contributes to nonhost resistance to inappropriate pathogens that enter by direct penetration. Plant Cell 18, 731–746, doi:10.1105/tpc.105.038372 (2006).
Kim, D. Y., Bovet, L., Maeshima, M., Martinoia, E. & Lee, Y. The ABC transporter AtPDR8 is a cadmium extrusion pump conferring heavy metal resistance. Plant J. 50, 207–218, doi:10.1111/j.1365-313X.2007.03044.x (2007).
Kühnlenz, T., Westphal, L., Schmidt, H., Scheel, D. & Clemens, S. Expression of Caenorhabditis elegans PCS in the AtPCS1-deficient Arabidopsis thaliana cad1-3 mutant separates the metal tolerance and nonhost resistance functions of phytochelatin synthases. Plant Cell Environ. 38, 2239–47, doi:10.1111/pce.12534 (2015).
Jozefczak, M. et al. Differential response of Arabidopsis leaves and roots to cadmium: Glutathione-related chelating capacity vs antioxidant capacity. Plant Physiol. Biochem. 83, 1–9, doi:10.1016/j.plaphy.2014.07.001 (2014).
Roth, U., Von Roepenack-Lahaye, E. & Clemens, S. Proteome changes in Arabidopsis thaliana roots upon exposure to Cd2+. J. Exp. Bot. 57, 4003–4013, doi:10.1093/jxb/erl170 (2006).
Villiers, F. et al. Investigating the plant response to cadmium exposure by proteomic and metabolomic approaches. Proteomics 11, 1650–1663, doi:10.1002/pmic.201000645 (2011).
Van De Mortel, J. E. et al. Expression differences for genes involved in lignin, glutathione and sulphate metabolism in response to cadmium in Arabidopsis thaliana and the related Zn/Cd-hyperaccumulator Thlaspi caerulescens. Plant Cell Environ. 31, 301–324, doi:10.1111/pce.2008.31.issue-3 (2008).
Chen, J. et al. Zinc-finger transcription factor ZAT6 positively regulates cadmium tolerance through the glutathione-dependent pathway in Arabidopsis. Plant Physiol. 171, 707–719, doi:10.1104/pp.15.01882 (2016).
Dutilleul, C., Jourdain, A., Bourguignon, J. & Hugouvieux, V. The Arabidopsis putative selenium-binding protein family: Expression study and characterization of SBP1 as a potential new player in cadmium detoxification processes. Plant Physiol. 147, 239–251, doi:10.1104/pp.107.114033 (2008).
Park, J. et al. The phytochelatin transporters AtABCC1 and AtABCC2 mediate tolerance to cadmium and mercury. Plant J. 69, 278–288, doi:10.1111/tpj.2011.69.issue-2 (2011).
Brunetti, P. et al. Cadmium-inducible expression of the ABC-type transporter AtABCC3 increases phytochelatin-mediated cadmium tolerance in Arabidopsis. J. Exp. Bot. 66, 3815–29, doi:10.1093/jxb/erv185 (2015).
Kühnlenz, T., Schmidt, H., Uraguchi, S. & Clemens, S. Arabidopsis thaliana phytochelatin synthase 2 is constitutively active in vivo and can rescue the growth defect of the PCS1-deficient cad1-3 mutant on Cd-contaminated soil. J. Exp. Bot. 65, 4241–4253, doi:10.1093/jxb/eru195 (2014).
van Ooijen, J. W. MapQTL® 5, Software for the mapping of quantitative trait loci in experimental population (2004).
Lander, E. S. & Botstein, D. Mapping mendelian factors underlying quantitative traits using RFLP linkage maps. Genetics 121, 185–199 (1989).
Zhu, T. et al. Toward elucidating the global gene expression patterns of developing Arabidopsis: Parallel analysis of 8300 genes by a high-density oligonucleotide probe array. Plant Physiol. Biochem. 39, 221–242, doi:10.1016/S0981-9428(00)01234-1 (2001).
R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org/ (2013).
Gentleman, R. C. et al. Bioconductor: open software development for computational biology and bioinformatics. Genome Biol. 5, R80, doi:10.1186/gb-2004-5-10-r80 (2004).
We thank Holger Schmidt for PC analyses, Michael Weber and Christiane Meinen for transcript analyses. This research was in part supported by the Deutsche Forschungsgemeinschaft (DFG) [FI 2152/1-1].
The authors declare that they have no competing interests.
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Fischer, S., Spielau, T. & Clemens, S. Natural variation in Arabidopsis thaliana Cd responses and the detection of quantitative trait loci affecting Cd tolerance. Sci Rep 7, 3693 (2017). https://doi.org/10.1038/s41598-017-03540-z
This article is cited by
Journal of Soil Science and Plant Nutrition (2023)
Genome-wide association study (GWAS) reveals genetic loci of lead (Pb) tolerance during seedling establishment in rapeseed (Brassica napus L.)
BMC Genomics (2020)